Rapid three-dimensional modeling method, medium and system based on seabed point cloud data

Abstract

The invention provides a rapid three-dimensional modeling method based on seabed point cloud data, a medium and a system, and belongs to the technical field of seabed depth measurement, and the method comprises the steps: firstly, obtaining low-resolution point cloud data through spatial downsampling, and carrying out the multivariate mixed decomposition of the low-resolution point cloud data, and obtaining a stable component and a variable component; constructing an initial three-dimensional grid model based on the stable component, and then dividing the initial three-dimensional grid model into a plurality of sub-regions by adopting a self-adaptive partitioning strategy; and performing partition processing on the original point cloud data according to sub-regions to obtain a plurality of high-resolution point cloud data blocks. And next, performing anomaly identification and noise reduction processing on each high-resolution point cloud data block, reconstructing a corrected high-resolution point cloud data block, performing grid reconstruction and boundary registration, and constructing a complete high-resolution seabed scene three-dimensional model. And the final high-quality seabed scene three-dimensional model is obtained through processing, so that the technical problem that the existing three-dimensional seabed model cannot fully reflect the detail characteristics of the submarine topography is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of seabed depth measurement, and in particular, relates to a fast three-dimensional modeling method, medium and system based on seabed point cloud data. Background Art

[0002] Seabed topography mapping is one of the basic tasks in the fields of marine scientific research and underwater engineering construction. With the continuous deepening of marine development activities, human demand for seabed topography information is also growing. Traditional seabed topography surveys usually rely on acoustic measurement technology, such as multi-beam echo sounders, single-beam echo sounders, etc. These technologies can obtain three-dimensional point cloud data of seabed topography, providing a basis for subsequent three-dimensional modeling and analysis. However, due to the complexity of the marine environment and the limitations of measurement equipment, these traditional methods still have certain limitations in data acquisition efficiency, resolution and accuracy. For example, the scanning range and resolution of acoustic measurement equipment are limited by many factors, such as water depth, ocean currents, turbulence, etc. In deep-sea environments, it is difficult for measurement equipment to obtain high-resolution seabed topography data. Therefore, the existing three-dimensional seabed models often cannot fully reflect the detailed characteristics of the seabed topography, and it is difficult to meet the needs of underwater engineering construction, marine resource exploration, etc. for high-precision seabed topography information. Summary of the invention

[0003] In view of this, the present invention provides a rapid three-dimensional modeling method, medium and system based on seabed point cloud data, which can solve the technical problem that existing three-dimensional seabed models often cannot fully reflect the detailed features of the seabed topography.

[0004] The present invention is achieved in that: The present invention provides a rapid three-dimensional modeling method based on seabed point cloud data, which includes the following steps: S01, collecting seabed point cloud data, and performing spatial downsampling processing on the seabed point cloud data to obtain low-resolution point cloud data; S02, performing multivariate mixed decomposition calculation on the low-resolution point cloud data to obtain low-resolution point cloud stable component data and low-resolution point cloud variable component data; S03, constructing an initial three-dimensional grid model of the seabed scene based on the low-resolution point cloud stable component data, and calculating the spatial curvature distribution of the initial three-dimensional grid model; S04, adaptively dividing the initial three-dimensional grid model into blocks according to the spatial curvature distribution, and dividing the initial three-dimensional grid model into a plurality of sub-regions; S05, spatially partitioning the original seabed point cloud data according to the multiple sub-regions to obtain multiple high-resolution point cloud data blocks; S06, performing multivariate mixed decomposition calculation on each of the high-resolution point cloud data blocks to obtain corresponding point cloud stable component data and point cloud variable component data respectively; S07, performing Fourier transform on the point cloud change component data of each high-resolution point cloud data block, establishing a point cloud spatial distribution feature matrix, extracting the main feature vector, and constructing a point cloud anomaly recognition model; S08, calculating the anomaly coefficient matrix of each of the high-resolution point cloud data blocks according to the point cloud anomaly recognition model, performing noise reduction processing on the point cloud change component data, and obtaining the point cloud correction component data; S09, combining and reconstructing the point cloud stabilization component data and the point cloud correction component data of each of the high-resolution point cloud data blocks to obtain a corrected high-resolution point cloud data block; S10, reconstructing a grid for each of the corrected high-resolution point cloud data blocks to generate a high-resolution sub-region three-dimensional model; S11, performing boundary registration and splicing on the high-resolution sub-region three-dimensional models to construct a complete high-resolution three-dimensional model of the seabed scene; S12, performing global geometry optimization and texture mapping on the high-resolution three-dimensional model of the seabed scene to obtain a final three-dimensional model of the seabed scene.

[0005] On the basis of the above technical solution, the rapid 3D modeling method based on seabed point cloud data of the present invention can also be improved as follows: Wherein, the step S01 specifically includes: collecting original point cloud data of the seabed area, and performing voxel downsampling processing on the original point cloud data; wherein, the voxel downsampling processing is to divide the three-dimensional space into a number of voxel units, perform statistical calculations on the point cloud data inside each voxel unit, and obtain the average value or median value of the points inside each voxel unit; the size of the voxel unit is adjustable within the range of 0.01 meters to 1 meter, and a balance between computational efficiency and the degree of retention of seabed terrain details is achieved by adjusting the size of the voxel unit.

[0006] Furthermore, the step S02 specifically includes: performing a multivariate mixed decomposition calculation on the low-resolution point cloud data, decomposing the low-resolution point cloud data into low-resolution point cloud stable component data and low-resolution point cloud variable component data; wherein the low-resolution point cloud stable component data represents the basic form of the seabed topography, and is linearly represented using 5 to 10 basis functions; the low-resolution point cloud variable component data represents slight changes in the seabed topography, and is obtained by subtracting the low-resolution point cloud stable component data from the low-resolution point cloud data.

[0007] The step S03 specifically includes: constructing an initial three-dimensional grid model of the seabed scene based on the low-resolution point cloud stable component data; calculating the spatial curvature distribution of the surface of the initial three-dimensional grid model, wherein the spatial curvature distribution is obtained by calculating the gradient operator and the Hessian matrix of each point of the initial three-dimensional grid model; the spatial curvature distribution is used to characterize the complexity and change characteristics of the seabed terrain.

[0008] The step S04 specifically includes: calculating the mean curvature and the standard deviation of curvature according to the spatial curvature distribution; taking the weighted sum of the mean curvature and the standard deviation of curvature as a blocking threshold, wherein the weight coefficient in the weighted sum ranges from 1 to 3; dividing the grid cells whose curvature values ​​are higher than the blocking threshold into separate sub-areas; and improving the geometric accuracy of the three-dimensional model in complex areas through the adaptive blocking strategy.

[0009] The step S05 specifically includes: spatially partitioning the original seabed point cloud data according to the sub-region division scheme of the initial three-dimensional grid model; dividing the partitioned original seabed point cloud data into multiple high-resolution point cloud data blocks; the high-resolution point cloud data blocks retain the complete spatial distribution information of the original seabed point cloud data.

[0010] The step S06 specifically includes: performing a multivariate mixed decomposition calculation on each of the high-resolution point cloud data blocks; using 10 to 20 detail basis functions to perform a linear decomposition on each of the high-resolution point cloud data blocks to obtain corresponding point cloud stable component data; and obtaining point cloud variable component data by subtracting the point cloud stable component data from the high-resolution point cloud data block.

[0011] The step S07 specifically includes: performing a three-dimensional discrete Fourier transform on the point cloud change component data of each of the high-resolution point cloud data blocks; establishing a point cloud spatial distribution feature matrix based on the results of the three-dimensional discrete Fourier transform; extracting a main eigenvector by performing feature analysis on the point cloud spatial distribution feature matrix; and constructing a point cloud anomaly recognition model based on the main eigenvector.

[0012] The step S08 specifically includes: calculating the anomaly coefficient matrix of each of the high-resolution point cloud data blocks according to the point cloud anomaly recognition model; selecting 3 to 5 main eigenvectors to construct the anomaly coefficient matrix; and using the anomaly coefficient matrix to perform denoising on the point cloud change component data to obtain the point cloud correction component data.

[0013] The step S09 specifically includes: linearly superimposing the point cloud stabilization component data and the point cloud correction component data of each of the high-resolution point cloud data blocks; obtaining a corrected high-resolution point cloud data block through the linear superposition; the corrected high-resolution point cloud data block has higher spatial accuracy and lower noise level.

[0014] The step S10 specifically includes: processing each of the corrected high-resolution point cloud data blocks using a point cloud meshing algorithm; the point cloud meshing algorithm includes a Poisson surface reconstruction algorithm or a spherical projection algorithm; and converting discrete point cloud data into a continuous three-dimensional mesh surface model through the point cloud meshing algorithm.

[0015] The step S11 specifically includes: extracting boundary point pairs of the high-resolution sub-region three-dimensional model; establishing a boundary registration error function, wherein the boundary registration error function includes a rigid body transformation term and a regularization term, and the coefficient of the regularization term ranges from 0.1 to 1; and achieving accurate splicing of each of the high-resolution sub-region three-dimensional models by minimizing the boundary registration error function.

[0016] The step S12 specifically includes: establishing a smoothing error function, wherein the smoothing error function includes a vertex displacement term and a normal vector term; a weight coefficient of the smoothing error function ranges from 0.01 to 0.1; geometrically optimizing the high-resolution seabed scene three-dimensional model by minimizing the smoothing error function; and performing texture mapping processing on the optimized three-dimensional model.

[0017] Furthermore, the spatial downsampling processing calculation process in step S01 is specifically expressed as follows: ; In the formula, is the original point cloud data set; is the downsampled point cloud data set; is the voxel downsampling function; is the voxel size, the value range is [0.01m, 1m]; is the downsampling error.

[0018] Furthermore, the multivariate mixed decomposition calculation in step S02 is specifically expressed as follows: ; ; ; In the formula, Stabilize the component for low-resolution point clouds; It is the variable component of low-resolution point cloud; For the basis functions; is the basis function coefficient; is the number of basis functions, generally 5-10.

[0019] Furthermore, the calculation of the spatial curvature distribution in step S03 is specifically expressed as follows: ; In the formula, For point The curvature value at ; is the implicit function of point cloud fitting; is the gradient operator; is the Hessian matrix; Calculate the error for the curvature.

[0020] Furthermore, the adaptive block calculation in step S04 is specifically expressed as follows: ; In the formula, is the block threshold; is the mean curvature; is the standard deviation of curvature; is the weight coefficient, and its value range is [1,3].

[0021] Furthermore, the high-resolution point cloud decomposition calculation in step S06 is specifically expressed as follows: ; ; ; In the formula, is the sub-region index; For the High-resolution point cloud data blocks; For stable weight; is the variable component; For the Detail basis functions; is the detail basis function coefficient; is the number of detail basis functions, generally 10-20.

[0022] Furthermore, the Fourier transform calculation in step S07 is specifically expressed as follows: ; In the formula, is the three-dimensional Fourier transform result; is the number of spatial sampling points; is an imaginary unit; is the frequency domain coordinate.

[0023] Furthermore, the calculation of the abnormal coefficient matrix in step S08 is specifically expressed as follows: ; ; In the formula, is the abnormal coefficient matrix; is the characteristic value; is the feature vector; The number of main features is usually 3-5; is the corrected change component.

[0024] Furthermore, the point cloud reconstruction calculation in step S09 is specifically expressed as follows: ; In the formula, It is the reconstructed high-resolution point cloud; is the reconstruction error.

[0025] Furthermore, the boundary registration calculation in step S11 is specifically expressed as follows: ; In the formula, is the registration error; is the number of sub-regions; For the The number of sub-region boundary points; is the rigid body transformation matrix; are corresponding point pairs; is the regularization coefficient, and its value range is [0.1,1]; is the identity matrix.

[0026] Furthermore, the geometric optimization calculation in step S12 is specifically expressed as follows: ; In the formula, is the smoothing error; is a set of vertices; is a collection of patches; is the Laplace operator; is the vertex displacement; is the patch normal vector; is the target normal vector; is the weight coefficient, and its value range is [0.01, 0.1].

[0027] A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions, and when the program instructions are run in a computer, they are used to execute the above-mentioned method for rapid three-dimensional modeling based on seabed point cloud data.

[0028] A third aspect of the present invention provides a rapid three-dimensional modeling system based on seabed point cloud data, which includes the above-mentioned computer-readable storage medium.

[0029] Compared with the prior art, the rapid three-dimensional modeling method, medium and system based on seabed point cloud data provided by the present invention have the following beneficial effects: 1. Improved geometric accuracy and resolution of the 3D seafloor model. By adaptively refining the original seafloor point cloud data, the detailed features of the complex seafloor topography can be effectively captured. Compared with the existing method based on acoustic measurement, the 3D model constructed by the method of the present invention can describe the seafloor topography more realistically and meticulously. This is of great significance for application scenarios such as underwater engineering design and resource exploration.

[0030] 2. Enhanced the ability of the 3D seafloor model to express dynamic changes. The method of the present invention can simultaneously depict the basic form and minor changes of the seafloor topography by decomposing the point cloud data into stable components and variable components. This ability is of great value for applications such as monitoring the evolution of seafloor topography and evaluating the impact of underwater engineering.

[0031] 3. Improved the automation and efficiency of 3D modeling. Compared with the traditional data processing process with more manual participation, the method of the present invention can more automatically extract high-quality 3D seabed models from the original point cloud data. This not only improves work efficiency, but also reduces subjective errors introduced by humans.

[0032] 4. Expanded the application scenarios of 3D seafloor models. With its high resolution and high precision, the 3D seafloor model constructed by the method of the present invention can provide more reliable basic data support for fields such as ocean exploration, underwater engineering, and resource development. This will promote the further development of related application technologies.

[0033] In general, the rapid three-dimensional modeling method based on seabed point cloud data proposed in the present invention has achieved remarkable technical progress in improving modeling efficiency, enhancing model expression capabilities, and expanding application scenarios, and has solved the technical problem that existing three-dimensional seabed models often cannot fully reflect the detailed characteristics of the seabed topography. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 A flow chart of the method provided by the present invention; Figure 2 It is a thermal map of curvature distribution of some areas of the seabed terrain in the embodiment; Figure 3 This is a graph showing the result of the frequency spectrum characteristic analysis of the point cloud variation component in the embodiment; Figure 4 It is a statistical distribution histogram of the sub-region boundary registration error in the embodiment. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0036] like Figure 1 As shown, it is a flow chart of a rapid three-dimensional modeling method based on seabed point cloud data provided by the first aspect of the present invention. The method comprises the following steps: S01, collecting seabed point cloud data, performing spatial downsampling processing on the seabed point cloud data, and obtaining low-resolution point cloud data; S02, performing multivariate mixed decomposition calculation on the low-resolution point cloud data to obtain low-resolution point cloud stable component data and low-resolution point cloud variable component data; S03. constructing an initial three-dimensional grid model of the seabed scene based on the low-resolution point cloud stable component data, and calculating the spatial curvature distribution of the initial three-dimensional grid model; S04, adaptively dividing the initial three-dimensional grid model into multiple sub-areas according to the spatial curvature distribution; S05, spatially partitioning the original seabed point cloud data according to multiple sub-regions to obtain multiple high-resolution point cloud data blocks; S06, performing multivariate mixed decomposition calculation on each high-resolution point cloud data block to obtain corresponding point cloud stable component data and point cloud variable component data respectively; S07, performing Fourier transform on the point cloud change component data of each high-resolution point cloud data block, establishing a point cloud spatial distribution feature matrix, extracting the main feature vector, and constructing a point cloud anomaly recognition model; S08, calculating the anomaly coefficient matrix of each high-resolution point cloud data block according to the point cloud anomaly recognition model, performing noise reduction processing on the point cloud change component data, and obtaining the point cloud correction component data; S09, combining and reconstructing the point cloud stabilization component data and the point cloud correction component data of each high-resolution point cloud data block to obtain a corrected high-resolution point cloud data block; S10, reconstructing the mesh of each corrected high-resolution point cloud data block to generate a high-resolution sub-region three-dimensional model; S11, performing boundary registration and splicing on the high-resolution sub-region 3D models to construct a complete high-resolution seabed scene 3D model; S12. Perform global geometry optimization and texture mapping on the high-resolution three-dimensional model of the seabed scene to obtain a final three-dimensional model of the seabed scene.

[0037] The specific implementation methods of the above steps are described in detail below: The specific implementation method of step S01 is to first collect the original point cloud data of the seabed area These raw point cloud data usually have high sampling density and resolution. In order to reduce the computational overhead of subsequent processing, it is necessary to Perform spatial downsampling. Specifically, the voxel downsampling method can be used. This method divides the three-dimensional space into a number of voxel units and performs statistical processing on the point cloud data inside each voxel, such as calculating the average or median value of the points inside the voxel, so as to obtain the low-resolution point cloud data after downsampling. . Voxel size Can be The downsampling process can be expressed as: ; in, is the voxel downsampling function, is the error introduced during the downsampling process.

[0038] The specific implementation of step S02 is to perform the low-resolution point cloud data obtained in step S01. Perform multivariate mixed decomposition calculations. This calculation process can convert low-resolution point cloud data Decomposed into two parts: stable component and variable components Among them, the stable component To represent the basic form of the seabed topography, a set of basis functions can be used To perform linear representation; variable components It describes the slight changes in the seafloor topography. The mathematical description is as follows: ; ; ; in, are basis function coefficients, is the number of basis functions, usually taken as The purpose of this step is to decompose the low-resolution point cloud data into stable components and variable components that can be processed independently.

[0039] The specific implementation of step S03 is to firstly stabilize the low-resolution point cloud component obtained in step S02. Construct an initial 3D mesh model of the seabed scene. In order to evaluate the geometric characteristics of the initial model, it is necessary to calculate the spatial curvature distribution of the model surface. The curvature is an important indicator of the complexity of the geometric shape and can be used to guide the subsequent adaptive mesh refinement. The curvature calculation formula is as follows: ; in, is the implicit function of point cloud fitting, is the gradient operator, is the Hessian matrix, is the error introduced in the curvature calculation process. By calculating the curvature value of each point in the initial 3D mesh model , we can better understand the complexity and changing characteristics of the seabed topography.

[0040] The specific implementation of step S04 is to calculate the spatial curvature distribution obtained in step S03. , the initial 3D mesh model is adaptively divided into blocks. The purpose is to subdivide the more complex areas and keep the lower mesh density in the flat areas. The specific block strategy is as follows: ; in, is the block threshold, is the mean value of the curvature, is the standard deviation of the curvature, is the weight coefficient, and its value range is When the curvature value of a grid cell exceeds the threshold When the area is , it is divided into a sub-area for subdivision. This adaptive block strategy can effectively improve the geometric accuracy of the 3D model in complex areas while maintaining a low computational overhead.

[0041] The specific implementation method of step S05 is to divide the original seabed point cloud data into sub-area division schemes according to the sub-area division scheme obtained in step S04. Perform spatial partitioning to obtain multiple high-resolution point cloud data blocks These high-resolution point cloud data blocks will be used for subsequent detailed modeling.

[0042] The specific implementation method of step S06 is to perform a Perform multivariate mixed decomposition calculations to obtain the corresponding stable components and variable components This step is similar to step S02, except that high-resolution point cloud data is processed here instead of the previous low-resolution point cloud data. The mathematical description is as follows: ; ; ; in, For the The detail basis functions, is the corresponding coefficient, is the number of detail basis functions, usually taken as The purpose of this step is to further separate the stable terrain information and change information in the high-resolution point cloud data, laying the foundation for subsequent point cloud anomaly detection and noise suppression.

[0043] The specific implementation method of step S07 is to calculate the change component of each high-resolution point cloud data block. Perform three-dimensional discrete Fourier transform (3D-DFT) to establish the point cloud spatial distribution feature matrix By analyzing the feature matrix , the main spectral feature vector can be extracted and a point cloud anomaly recognition model can be constructed. The mathematical description is as follows: ; in, is the number of spatial sampling points, is an imaginary unit, is the frequency domain coordinate. By performing feature analysis, the main spectral feature vectors can be extracted, which will form the basis of the point cloud anomaly recognition model.

[0044] The specific implementation of step S08 is to calculate the abnormal coefficient matrix of each high-resolution point cloud data block according to the point cloud abnormality recognition model established in step S07 . Then use this abnormal coefficient matrix Change component of point cloud Perform noise reduction to obtain the corrected point cloud change component The mathematical description is as follows: ; ; in, is the characteristic value, is the feature vector, is the number of main eigenvectors, usually taken as Through this abnormal point cloud suppression method based on principal component analysis, the noise and interference components in high-resolution point cloud data can be effectively reduced.

[0045] The specific implementation of step S09 is to convert the stable component of each high-resolution point cloud data block into and the corrected change component Perform combined reconstruction to obtain the final corrected high-resolution point cloud data block The mathematical description is as follows: ; in, The error introduced in the reconstruction process. The purpose of this step is to fuse the stable component obtained in the previous step with the corrected variable component to obtain the final high-quality point cloud data block.

[0046] The specific implementation method of step S10 is to perform a correction on each high-resolution point cloud data block. Mesh reconstruction is performed to generate the corresponding high-resolution sub-region 3D model. This step can use various common point cloud meshing algorithms, such as Poisson surface reconstruction, spherical projection, etc. Through this mesh reconstruction process, discrete point cloud data can be converted into a continuous 3D mesh surface model.

[0047] The specific implementation of step S11 is to perform boundary registration and splicing of the high-resolution sub-region 3D models generated in step S10, so as to construct a complete high-resolution seabed scene 3D model. The mathematical description of boundary registration is as follows: ; in, is the registration error, is the number of sub-regions, For the The number of boundary points of each sub-region, is the rigid body transformation matrix, are the corresponding boundary point pairs, is the regularization coefficient, and its value range is , is the unit matrix. By minimizing the registration error function , which can achieve accurate splicing between the 3D models of each sub-area.

[0048] The specific implementation of step S12 is to perform global geometry optimization and texture mapping processing on the high-resolution seabed scene three-dimensional model constructed in step S11 to further improve the visual quality of the model. The mathematical description of geometry optimization is as follows: ; in, is the smoothing error, is a set of vertices, is a collection of patches, is the Laplace operator, is the vertex displacement, is the patch normal vector, is the target normal vector, is the weight coefficient, and its value range is By minimizing the smooth error function , the geometric shape of the 3D model can be optimized to make it smoother and more natural. At the same time, realistic surface details can be added to the 3D model through texture mapping and other technologies to further enhance the visual effect.

[0049] In summary, the fast 3D modeling method based on seabed point cloud data proposed in this invention involves multiple mathematical formulas and algorithm implementation details. This method can effectively extract the stable characteristics and dynamic change information of the seabed topography from the original high-density point cloud data, and finally construct a high-quality 3D seabed scene model through technical means such as adaptive mesh refinement, point cloud anomaly detection and geometric optimization.

[0050] A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions, and when the program instructions are run in a computer, they are used to execute the above-mentioned method for rapid three-dimensional modeling based on seabed point cloud data.

[0051] A third aspect of the present invention provides a rapid three-dimensional modeling system based on seabed point cloud data, which includes the above-mentioned computer-readable storage medium.

[0052] Specifically, the principle of the present invention is to effectively process and analyze the original high-resolution seabed point cloud data, extract stable terrain features and dynamic change information from it, and finally construct a high-quality three-dimensional seabed scene model. The method mainly includes the following key technical links: 1. Spatial downsampling. Since the original seabed point cloud data is usually large and dense, direct 3D modeling will face huge computational overhead. Therefore, the present invention first performs spatial downsampling on the original point cloud data to obtain low-resolution point cloud data. This step can greatly reduce the computational complexity of subsequent processing while retaining key terrain features.

[0053] 2. Point cloud decomposition. The downsampled low-resolution point cloud data is decomposed by multivariate hybrid decomposition to separate stable terrain features (stable components) and dynamic change information (variable components). This decomposition can provide a favorable data basis for subsequent adaptive mesh refinement and abnormal point cloud detection.

[0054] 3. Adaptive mesh refinement. Based on the stable components of the low-resolution point cloud, an initial 3D mesh model is first constructed. Then, according to the spatial curvature distribution of the mesh cells, the mesh is locally refined using an adaptive blocking strategy. This ensures overall computational efficiency while providing a more refined geometric expression for complex terrain areas.

[0055] 4. Frequency domain analysis and anomaly detection. For each high-resolution point cloud data block, the 3D discrete Fourier transform is used to extract its spectral features. By analyzing these spectral features, a point cloud anomaly recognition model can be constructed to effectively suppress the noise and interference components in the original point cloud data.

[0056] 5. Point cloud reconstruction and mesh generation. The stable components and the variable components detected by anomalies are combined and reconstructed to obtain the final high-quality point cloud data. Based on these point cloud data, the corresponding 3D model is generated by mesh reconstruction. At the same time, a complete high-resolution seabed scene model can be constructed by performing boundary registration and splicing on the sub-region 3D models.

[0057] 6. Global optimization and texture mapping. To further enhance the visual effect of the 3D seabed model, global geometry optimization and texture mapping can be performed on it. This includes optimizing the model geometry using methods such as Laplace smoothing, and enhancing the surface details of the model through texture mapping.

[0058] Through the organic combination of the above technical links, the method of the present invention can efficiently extract stable terrain features and dynamic change information from the original seabed point cloud data, and construct a high-resolution, high-precision three-dimensional seabed scene model. This method makes full use of the inherent characteristics of point cloud data, and overcomes the shortcomings of existing three-dimensional modeling methods based on acoustic measurement in complex terrain modeling and dynamic change expression through means such as adaptive grid refinement and frequency domain analysis.

[0059] It should be noted that the explanation of relevant variables in this specific implementation is shown in Table 1 below: Table 1 Variable explanation table

[0060] The following is an example of a specific application scenario of the present invention: Taking the seabed topography mapping task in a certain sea area as an example, the specific implementation process of the method of the present invention is introduced. The sea area is located on the southeast coast of China, with a water depth ranging from 50 to 1000 meters. The terrain is complex and changeable, including steep submarine mountains and gentle continental shelf areas. In order to obtain a high-precision three-dimensional seabed model of the sea area, the marine department decided to use the fast three-dimensional modeling method based on seabed point cloud data proposed in the present invention for mapping.

[0061] First, the surveying and mapping team used a multi-beam echo sounder to collect comprehensive seabed topography data in the sea area. After several days of continuous operation, a total of about 250 million seabed point cloud data were obtained. Since the acquisition density of the original point cloud data is very high, it will generate huge computing overhead in the subsequent 3D modeling process. To this end, the surveying and mapping team first collected the original point cloud data. Spatial downsampling was performed.

[0062] Specifically, the surveying and mapping team divided the entire sea area into a 1mx1mx1m voxel grid, and statistically processed the point cloud data inside each voxel to extract the average coordinate value of the voxel center position. The purpose of this is to significantly reduce the computational complexity of subsequent processing while retaining the key features of the seabed topography. After this step, the original 250 million point cloud data was compressed into about 30 million low-resolution point cloud data. Table 2 gives the comparison between the original point cloud data and the downsampled point cloud data.

[0063] Table 2 Point cloud data downsampling results index Original point cloud data Point cloud data after downsampling Point cloud quantity 250 million 30 million Average point spacing 0.2 m 1 m Computational complexity Very high Lower Next, the surveying and mapping team downsampled the low-resolution point cloud data. Multivariate mixed decomposition calculations were performed. Through this step, the point cloud data can be decomposed into point cloud stable components that reflect the stable characteristics of the seabed topography. , and the point cloud change component describing the dynamic changes of the seafloor topography .

[0064] Specifically, the surveying and mapping team chose to use 5 basis functions To fit the stable component of the point cloud . Solve the basis function coefficients by the least squares method , we can get the following mathematical expression: ; ; Among them, the basis function The choice is based on the low-resolution point cloud data It is determined by the analysis of the point cloud and can better express the basic form of the seabed topography. The undulations reflect the subtle fluctuations relative to the stable terrain. Through this decomposition method, the surveying and mapping team obtained the stable characteristics and dynamic information of the seabed terrain.

[0065] Based on point cloud stabilization component The mapping team constructed an initial 3D mesh model. To evaluate the geometric properties of the mesh model, the team calculated the spatial curvature of each mesh cell. The larger the curvature value, the more complex the terrain in the area. By analyzing the global curvature distribution, the team found that there are many steep terrain features in the sea area, such as submarine mountains and deep-water trenches.

[0066] like Figure 2The figure shows the curvature distribution heat map of some areas of the seabed terrain. The color depth shows the distribution of curvature values ​​in different areas, which helps to identify complex terrain areas.

[0067] In order to better depict these complex terrain areas, the surveying and mapping team adopted an adaptive mesh refinement strategy. Specifically, the average curvature of the initial mesh model is first calculated. and curvature standard deviation , and then set an adaptive block threshold For those grid cells with curvature values ​​above the threshold, the team divided them into finer sub-regions for further processing. Through this adaptive refinement method, a high-resolution grid model consisting of 30,000 sub-regions was finally obtained.

[0068] Next, the surveying team targeted each high-resolution sub-area point cloud data block. Further processing was performed. First, the surveying team used 10 detail basis functions Point cloud data block Perform multivariate mixture decomposition to obtain stable components and variable components : ; ; ; Among them, the detail basis function The choice of takes into account the topographic characteristics of the sub-region and can describe the local topographic changes more finely.

[0069] For each sub-region, the point cloud variation component The surveying and mapping team further used 3D discrete Fourier transform to extract its spectral features. By analyzing these spectral features, the team found that most of the high-frequency components were related to the noise and interference of the seabed topography. Therefore, the surveying and mapping team constructed an abnormal point cloud recognition model based on principal component analysis to identify abnormal point clouds. Perform effective noise reduction.

[0070] Specifically, the team extracted the first three main spectral feature vectors and its corresponding eigenvalues , the following abnormal coefficient matrix is ​​constructed : ; ; in, Represents the change component of the point cloud after noise suppression.

[0071] like Figure 3 As shown, the results of spectral feature analysis of the point cloud variation component are shown, showing the spectral distribution of the three main eigenvectors.

[0072] With the processed stable components of each sub-region and variable components The surveying and mapping team reconstructed and combined them to obtain the final high-resolution point cloud data block. : ; Based on these high-resolution point cloud data, the surveying and mapping team used the Poisson surface reconstruction algorithm to grid each sub-area and generate the corresponding high-resolution 3D mesh model. In order to splice these sub-area models into a complete seabed scene model, the team also implemented a boundary registration step. Specifically, the surveying and mapping team minimized the following registration error function To optimize the geometric consistency between the sub-region models: ; in, For the The rigid body transformation matrix of the sub-region model, Through this registration optimization, the surveying and mapping team finally accurately spliced ​​the 30,000 sub-area models into a complete high-resolution 3D model of the seabed scene.

[0073] like Figure 4 As shown, it is a statistical distribution histogram of the sub-region boundary registration error, which shows the error distribution during the registration process.

[0074] In order to further improve the visual effect of the 3D seafloor model, the surveying and mapping team also implemented global geometry optimization and texture mapping processing. Specifically, the surveying and mapping team used the Laplace smoothing algorithm to optimize the geometry of the model and minimize the following smoothing error function: : ; in, is the Laplace operator, is the vertex displacement, are the actual patch normal vector and the target normal vector respectively. By minimizing the error function, the geometric shape of the overall model can be made smoother and more natural.

[0075] Finally, the surveying and mapping team used high-resolution aerial images to add realistic textures to the 3D model. This not only improved the visual effect of the model, but also further enhanced its ability to restore the actual seabed environment. Table 3 lists the main parameters of the 3D model construction process of this sea area.

[0076] Table 3 Main parameters of the 3D seafloor model construction process

[0077] Through the above steps, the surveying and mapping team finally built a high-precision 3D seabed scene model covering the entire surveyed sea area with a resolution of 1 m. This model not only accurately depicts the complex seabed terrain features, such as steep seabed mountains and deep and wide trenches, but also can effectively reflect the dynamic changes of the seabed terrain.

[0078] This three-dimensional seabed model not only provides reliable basic data support for subsequent marine scientific research, but also brings important reference value to related underwater engineering construction projects. For example, the marine department can use this model to simulate and optimize the underwater pipeline laying plan, and the water resources exploration department can also carry out accurate exploration activities based on this model. In short, the fast three-dimensional modeling method based on seabed point cloud data proposed in this invention brings new technical support to various applications in the marine field.

[0079] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A rapid three-dimensional modeling method based on seabed point cloud data, characterized in that: The following steps are involved: S01, collecting seabed point cloud data, and performing spatial downsampling processing on the seabed point cloud data to obtain low-resolution point cloud data; S02, performing multivariate mixed decomposition calculation on the low-resolution point cloud data to obtain low-resolution point cloud stable component data and low-resolution point cloud variable component data; S03, constructing an initial three-dimensional grid model of the seabed scene based on the low-resolution point cloud stable component data, and calculating the spatial curvature distribution of the initial three-dimensional grid model; S04, adaptively dividing the initial three-dimensional grid model into blocks according to the spatial curvature distribution, and dividing the initial three-dimensional grid model into a plurality of sub-regions; S05, spatially partitioning the original seabed point cloud data according to the multiple sub-regions to obtain multiple high-resolution point cloud data blocks; S06, performing multivariate mixed decomposition calculation on each of the high-resolution point cloud data blocks to obtain corresponding point cloud stable component data and point cloud variable component data respectively; S07, performing Fourier transform on the point cloud change component data of each high-resolution point cloud data block, establishing a point cloud spatial distribution feature matrix, extracting the main feature vector, and constructing a point cloud anomaly recognition model; S08, calculating the anomaly coefficient matrix of each of the high-resolution point cloud data blocks according to the point cloud anomaly recognition model, performing noise reduction processing on the point cloud change component data, and obtaining the point cloud correction component data; S09, combining and reconstructing the point cloud stabilization component data and the point cloud correction component data of each of the high-resolution point cloud data blocks to obtain a corrected high-resolution point cloud data block; S10, reconstructing a grid for each of the corrected high-resolution point cloud data blocks to generate a high-resolution sub-region three-dimensional model; S11, performing boundary registration and splicing on the high-resolution sub-region three-dimensional models to construct a complete high-resolution three-dimensional model of the seabed scene; S12, performing global geometry optimization and texture mapping on the high-resolution three-dimensional model of the seabed scene to obtain a final three-dimensional model of the seabed scene.

2. A rapid three-dimensional modeling method based on seabed point cloud data according to claim 1, characterized in that: The step S01 specifically includes: collecting original point cloud data of the seabed area, and performing voxel downsampling processing on the original point cloud data; wherein the voxel downsampling processing is to divide the three-dimensional space into a number of voxel units, perform statistical calculations on the point cloud data inside each voxel unit, and obtain the average value or median value of the points inside each voxel unit; the size of the voxel unit is adjustable within the range of 0.01 meters to 1 meter, and a balance between computational efficiency and the degree of retention of seabed terrain details is achieved by adjusting the size of the voxel unit.

3. The rapid three-dimensional modeling method based on seabed point cloud data according to claim 2 is characterized in that: The step S02 specifically includes: performing a multivariate mixed decomposition calculation on the low-resolution point cloud data, decomposing the low-resolution point cloud data into low-resolution point cloud stable component data and low-resolution point cloud variable component data; wherein the low-resolution point cloud stable component data represents the basic form of the seabed topography, and is linearly represented using 5 to 10 basis functions; the low-resolution point cloud variable component data represents slight changes in the seabed topography, and is obtained by subtracting the low-resolution point cloud stable component data from the low-resolution point cloud data.

4. The rapid three-dimensional modeling method based on seabed point cloud data according to claim 3 is characterized in that: The step S06 specifically includes: performing a multivariate mixed decomposition calculation on each of the high-resolution point cloud data blocks; using 10 to 20 detail basis functions to perform a linear decomposition on each of the high-resolution point cloud data blocks to obtain corresponding point cloud stable component data; and obtaining point cloud variable component data by subtracting the point cloud stable component data from the high-resolution point cloud data block.

5. The rapid three-dimensional modeling method based on seabed point cloud data according to claim 4 is characterized in that: The step S07 specifically includes: performing a three-dimensional discrete Fourier transform on the point cloud change component data of each of the high-resolution point cloud data blocks; establishing a point cloud spatial distribution feature matrix based on the results of the three-dimensional discrete Fourier transform; extracting a main eigenvector by performing feature analysis on the point cloud spatial distribution feature matrix; and constructing a point cloud anomaly recognition model based on the main eigenvector.

6. The rapid three-dimensional modeling method based on seabed point cloud data according to claim 5, characterized in that: The step S09 specifically includes: linearly superimposing the point cloud stabilization component data and the point cloud correction component data of each of the high-resolution point cloud data blocks; obtaining a corrected high-resolution point cloud data block through the linear superposition; the corrected high-resolution point cloud data block has higher spatial accuracy and lower noise level.

7. The rapid three-dimensional modeling method based on seabed point cloud data according to claim 6, characterized in that: The step S11 specifically includes: extracting boundary point pairs of the high-resolution sub-region three-dimensional model; establishing a boundary registration error function, wherein the boundary registration error function includes a rigid body transformation term and a regularization term, and the coefficient of the regularization term ranges from 0.1 to 1; and achieving accurate splicing of each of the high-resolution sub-region three-dimensional models by minimizing the boundary registration error function.

8. The rapid three-dimensional modeling method based on seabed point cloud data according to claim 7, characterized in that: The step S12 specifically includes: establishing a smoothing error function, wherein the smoothing error function includes a vertex displacement term and a normal vector term; a weight coefficient of the smoothing error function ranges from 0.01 to 0.1; geometrically optimizing the high-resolution seabed scene three-dimensional model by minimizing the smoothing error function; and performing texture mapping processing on the optimized three-dimensional model.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program instructions, and when the program instructions are executed in a computer, they are used to execute the rapid three-dimensional modeling method based on seabed point cloud data as described in any one of claims 1 to 8.

10. A rapid three-dimensional modeling system based on seabed point cloud data, characterized in that: A computer-readable storage medium comprising the computer-readable storage medium of claim 9.

Images